Several readers have corresponded with me recently about my series of columns on North Carolina’s recent drop in unemployment. Why, they ask, am I devoting so much of my attention to this issue?
There are two reasons. First, the performance of North Carolina’s labor market over the past year is of great importance. The economy is the number-one concern of most North Carolinians, be they regular voters or political bystanders.
The second reason I have devoted and will continue to devote attention to North Carolina’s recent drop in unemployment is that, to be blunt, journalists and policy analysts have as a group failed to provide their audiences with accurate information on the matter. For some, this simply reflects unfamiliarity with the topic. For others, I’m afraid that the explanation is far less benign.
Every month, the U.S. Bureau of Labor Statistics conducts and releases two separate surveys about the labor market. One of them, a fairly large survey of employers, generates counts of how many people are on payrolls, among other things. When analysts talk about the number of “jobs” created or lost, this is the dataset to which they are usually referring.
The other, smaller “household survey” of individuals produces the unemployment rate you hear about every month, along with counts of how many people are employed, how many people are unemployed but looking for work, and the size of the civilian labor force (which is nothing more than the sum of the first two counts).
Over time, the two surveys tend to track together, although keep in mind that the number of payroll jobs is typically lower than the number of employed people (because of start-up firms that don’t immediately show up in the payroll survey, for example). On a month-to-month basis, however, the two trend lines can diverge. Over short periods of time, the larger-sample payroll survey is less likely to be erroneous. Over long periods of time, however, the household survey provides a much richer set of data about the labor market, including what jobless workers are doing if they’re not actively looking for work.
At the state level, BLS produces annual averages for labor-market measures that are much broader than the standard U-3 unemployment rate. For example, the bureau’s U-4 rate includes discouraged workers who’ve given up looking for a job and are no longer counted in the labor force. The U-5 rate includes those workers as well as others who’ve stopped looking for work for any other reason, such as those returning to school to train for a new profession. Finally, the U-6 rate adds in workers who’ve taken part-time jobs but would still like to work full-time, a phenomenon often called “underemployment.”
From June 2013 to July 2013, North Carolina experienced one of the largest drops in U-3 unemployment in the country. Although you’ll see journalists or politicians claim this drop was “largely,” “significantly,” or “in the main” caused by jobless North Carolinians leaving the labor force, such claims are unequivocally false. The vast majority of the drop represents employment gains, not declines in the labor force.*
Now that the BLS has updated its broader labor-market measures, there is even more relevant evidence. During the 12 months that ended in June 2013, North Carolina’s U-5 rate (including all jobless workers not currently seeking employment) averaged 10.5 percent. A year later, through June 2014, North Carolina’s U-5 rate averaged 8.1 percent. That drop of 2.4 percentage points was the second largest in the country, and far larger than the national average (-1.1) or regional average (-1.3 to -1.6, depending on how you define our region).
North Carolina’s U-6 rate, which treats involuntary part-timers as less than fully employed, was 13 percent through June 2014, down from 15.6 percent in June 2013. North Carolina’s 2.6-point drop was a tie for fourth largest, and again much higher than the national and regional averages.
It is entirely legitimate to disagree about why North Carolina’s labor market has improved so dramatically over the past year. But denying the reality of the improvement is possible only if you are ignorant of the data or deeply dishonest.
John Hood is president of the John Locke Foundation.
*By the way, perhaps the strangest claim people make about North Carolina’s exit from unemployment-insurance extended benefits in July 2013 is that it “caused” jobless North Carolinians to drop out of the labor force, thus artificially decreasing the state’s measured unemployment rate. While it is true that jobless recipients have to tell the state they are looking for work in order to be eligible for UI checks, that has nothing to do with the household survey conducted by the Bureau of Labor Statistics. If a UI recipient happens to end up in the confidential BLS sample, he can tell the pollster he’s not really looking for work if that is, in fact, true (see Constanza, George). It will have no effect on his UI checks. And if someone who exhausts his UI benefits or was never eligible in the first place (which represents most unemployed people, by the way) gets in the BLS sample, he can say he’s looking for work all he wants — it still won’t get him a UI check.